Using facial features to detect autism is a growing area of research. While it can aid in diagnosis, much of the research remains unreliable.

If you live with autism spectrum disorder (ASD) or suspect you have autism, getting a diagnosis can be difficult. Researchers and clinicians are constantly looking for ways to diagnose conditions like autism more effectively and accurately.
People with autism may have facial features that are different from the average person. If autism is diagnosed by looking at an individual's facial features, this can help speed up diagnosis and get the support they need sooner.
Do autistic people have distinctive facial features?
Yes, some research supports the idea that people with autism have a specific set of facial features. A 2019 study of autistic children found two facial signs that help identify autism; a reduced height of the midline of the face and wide-set eyes.
This study had a small sample size and was conducted only with white children. Due to the limitations of the study, it lacks reliability.
Additional research on facial features in children with autism supports the findings of the first study, suggesting that facial features of autism include:
- The upper face is wider
- Larger mouth and philtrum
- Shorter mid face
- Wider eyes
The authors also suggest that differences in the brain development of people with autism may produce different facial features. Using physical cues to help detect autism is a growing area of research.
It's important to note that because research in this area is limited, seeing these facial features does not mean you have autism. You can still have autism even if you don't have them.
Diagnosis of autism in children today is often based on observational tools that assess social and cognitive development. There are currently two main tools that help diagnose autism:
- Diagnostic and Statistical Manual of Mental Disorders, 5th Edition, Text Revision (DSM-5-TR) : outlines the criteria that many mental health professionals use to diagnose autism spectrum disorder
- Modified Checklist for Autism in Toddlers (M-CHAT) : a checklist of developmental symptoms present or absent used to help diagnose children aged 16-30 months.
Can facial features help diagnose autism?
The answer is yes. However, they should not be relied upon solely to diagnose autism. A 2022 study looked at several models that aimed to detect autism using facial features and found that many of them could detect autism with 86%–95% accuracy. The researchers suggest that computational tools that can detect faces could help doctors make diagnoses.

Physical characteristics of people with autism
There is little or no research into other physical features of autism that help detect the condition. There are notable features that may prompt parents to have their child evaluated for autism at an early age.
Early signs of autism include:
- Delayed language development, such as understanding speech and pronunciation
- Difficulty in movement
- Poor eye contact
- Social difficulties, such as not being interested in other children, not responding when called by name
- Restricted or repetitive interests
- Difficult to change habits
- Trouble with nonverbal body language, such as pointing or gesturing